342 ALASKAN GLACIER STUDIES side has long sloping spurs not truncated excepting at the tip, nor oversteepened so much as the eastern side. This might possibly be explained by a difference in the rock,1 for part of the eastern shore is granite, while the western coast is all slate and graywacke, were it not for the fact that southern Esther Island arid the coast north of Esther Passage, on the eastern shore, are made up of slate and graywacke, and yet these parts are nearly as much oversteepened as the granite coast. The probable explanation is that there has been differential glacial erosion on the opposite sides of the fiord, the former Port Wells Glacier probably hugging the eastern shore because many tributaries entered from the opposite side. Cirques and, Hanging Valleys. Another evidence of profound glacial erosion comes from the cirques and hanging Valleys, of which there are many. There is a well-developed hanging valley, for example, on the western side of Barry Arm, about a mile south of Cascade Glacier (PI. CXXXEX). The upper part of this valley is still occupied by a small glacier from which a stream descends to the fiord, first with a moderate grade, then precipitously over the lip of the hanging valley which lies between 1000 and 1100 feet above sea level. The stream cascades down to Barry Arm in a large waterfall and a filmy bridal-veil fall, plunging nearly 500 feet in one vertical cascade and altogether descending 1050 feet with horizontal flow of less than 500 feet. This hanging valley has only recently been revealed and the cascades existed only a few years, for in 1898 the Barry Glacier extended up to the lip of the hanging valley. The hanging condition was partly revealed in 1899 but not fully until some time between 1905 and 1908 when the terminus of Barry Glacier had retreated northeast of the site of the waterfall so that the full height of the hanging valley above sea level was exposed. Besides this typical hanging valley there are many others, of which those containing Cascade, Cataract and Toboggan Glaciers, and the small ice tongues southeast of Barry Glacier may be cited as examples. There are also many cirques from which the ice is completely melted, as between Cascade and Serpentine Glaciers, and on the slopes of Mt, Muir. Glacial Erosion below Sea Level—Longitudinal Profiles of Fiords. The depth of water at the fronts of the tidal glaciers (Barry 426 feet, Surprise 854 feet, Cataract 150 feet, Harriman 90 feet, Serpentine 51 feet) proves that these ice tongues are all resting on the bottom of the fiord and are therefore still engaged in eroding it. Within Harriman Fiord the depth of water varies from 90 to 510 feet (Fig. 50), and the longitudinal profiles are undulating, due either to glacial basining or to complication of glacial deposits, a point discussed below. The fiord bottom slope, from Pt. Doran to Pt. Pakenham, increasing evenly in depth from 414 to 576 feet, averages 82 feet to the mile, and tTn'a slope is not interrupted by basins due to erosion, or by note worthy submerged glacial deposits. At Pt. Pakenham where the fiord is shallowed locally by what is interpreted as a submerged moraine, the depth increases suddenly from 576 to 980 feet (PI. CXLI). As this is at the junction of Barry Arm and College Fiord we infer that the former expanded Port Wells glacier, which was fed from these inlets, was able to erode more efficiently than either the smaller Barry Arm or the College Fiord glaciers, but as they were sub-equal in size they have produced a confluence step at the north end of Port Wells. The longitudinal profile is interrupted by a step at the mouth of each fiord, however, > See geological map by Grant and Biggins, PI. n. Bull. 443, U. S. Geol. Survey, 1910.